Gold put in quite a negative day yesterday, its 9th worst dating back to 1975. Should we buy the dip or wait and see?

Gold had quite a day, and not for the same reasons we have been talking about for the last few weeks. The SPDR Gold Trust (GLD) slipped nearly 6.5% on 10/21/2025, its worst day since a nearly 9% daily decline in April of 2013. The move brought GLD back into a column of O’s on its default chart, retreating quickly off all-time highs. Despite this, we are still heavily extended (weekly OBOS of +150%) without notable support on our default chart. As with any large move, it begs the question: Is this the start of something serious or maybe just an opportunity to garner further exposure to an otherwise strong asset class?

Of the roughly 13,250 trading days since 1975, this move is the ninth worst exhale and one of only 13 that saw GLD down more than 6% on a single day. The table below details other notable days, defined by a >6% fall on a single day (do note, clusters are not excluded in our dataset below). Large declines clustered around the late 1970’s and early 1980 (rate hikes to combat aggressive inflation) but other notable declines came in 2006, 2008, and previously mentioned 2013. Not excluding these clusters, forward returns are muted, seeing the one year average return hover around +.1%. That fact alone does at least offer some credence to the idea that it is time for the metal to slow down. After all, GLD is still up ~55% this year, its second best showing through 10/21 of a calendar year since the start of our data set in 1975, only following the 73%+ gain through 10/21 of 1979. Out of curiosity, we included the admittedly limited GLD fund score data (comparatively to this dataset) to see if that might provide more context. The table includes GLD fund scores for large single day declines in 2013 (1.49), 2006 (4.69) and 2008 (5.46). As you might expect, longer-term forward performance metrics are more attractive when fund scores are strong. This fact, despite offering a small dataset… offers some compelling fuel to the idea that this is nothing more than a pullback for an otherwise strong asset class.

With that idea in mind, we can continue to zoom out to provide more context to the recent exhale. The distribution table below breaks down rolling 20-day (roughly one month, including weekends) performance metrics for GLD. Before 10/21, that 20-day advance stood at roughly 16.5%, a 99 percentile reading. Even with the decline on 10/21, the fund is still up just under 10% over the last respective 20 trading days. This represents a 96.3 percentile reading. All this to say… yes, the decline is certainly notable but recent performance has been so strong that the daily decline is a drop in the bucket. On the other hand, a more pessimistic view would suggest that there is still plenty of room to run to the downside before a trading range (or rolling performance range for that matter) is “normal” by any form of the word.

There are two trains of thought here:

• Buy strong assets relentlessly, take pullbacks as an opportunity to double down.
• Wait and see if more consolidation is in store, avoid selling positions until further weakness is established but avoid adding more to current holdings.

Which makes sense to you is largely going to be dependent on your situation. If you already have an oversized position within gold focused names, play more of a wait and see game. For those of you who had waited on the sidelines while gold ran in your face, take the decline as an opportunity to add lightly, with the understanding that there could be more back and forth in store. The empirical evidence behind trend following would suggest you simply buy strong assets without thinking twice, but doing so isn’t always appropriate from a business perspective. Regardless of your next step forward, continue to monitor the space as we move into November.